Effect of Phosphate Depletion on Magnesium

Homeostasis in Rats

WILHELMJ. KREUSSER,KIYoSHI KUROKAWA,ENRIQUEAZNAR, ELLEN SACHTJEN, andSHAULG. NMASSRY,Division of Nephrology and the Department of Medicinie,University of Southern Californiia School of Medicitne, Los Atngeles, Californiia90033

A B S T R A C T The effects of phosphate depletion onmagniesiumni (NMg) homeostasis were evaluated in ratsfed a diet containing 0.03% phosphorus for periods upto 8 wk. Plasma phosphorus fell signiificantly (P < 0.01)from 10.1+0.27 (SE) to 5.0±0.54 mgIl00 ml within 1day and conitiniued to fall gradually to a level of 1.2±0.21mg/100 ml by the end of the 8th wk. A signiificant(P < 0.01) increment in urinary Mg excretion (UMgV)from 46±2.7 to 126±24 ,ueq/24 h occurred during the1st day of phosphate depletion; UMgVreached a peak of300±+24 pieq/24 h by the 3rd day and remained highranging between 150-300 ,ueq/24 h, thereafter. Themagnittude of the magnesuria was related to the degreeof hypophosphatemia and was not affected by loweringthe calcium intake and reducing the hypercalciuria.The concentration of plasma Mg fell significantly(P < 0.01) from 1.2±0.02 to 0.79±0.10 meq/liter by the1st day of the study and remained low throughout.

Mg balance became negative during the 1st day ofphosphate depletion and remained so during the entirestudy. This occurred despite a significant increment inthe fraction of ingested Mg absorbed which becarmieevident by the 3rd wk of phosphate depletion. Mgcontenit of muscle, kidney, and liver were not affectedbut bone Mgwas reduced significantly. The change inbone Mg was not due to an overall reduction in bonemineral contenit because bone calcium content was notaffected. Supplementation of large amounts of Mg(800-1,000 ,ueq/day) in the drinking water produced anormalization of serum Mg but did not bring aboutrestoration of bone Mg despite a positive Mgbalance.The disturbances in Mgmetabolism were independentof the age or veight of the animals.

Our results indicate that phosphate depletion is

Dr. Kreusser is a Fellow of the Deutsche Forsch-ungsgemeinschaft.

Dr. Aznar is a Fellow of the Del Amo Foundation.Received for publicationi 10 AMay 1977 and in revised form

27 October 1977.

associated with (a ) magniesuria due to a decrease in thenet renal tubular reabsorption of Mg with the mainsource of the urinary losses being bone Mg; (b)hypomagnesemia secondary to the renal leak of Mg; (c)negative Mgbalance; and (d) increase in the intestinalfractional absorption of Mg. The latter was not adequateto compensate for the urinary losses of Mg.

INTRODUCTION

Phosphate depletion in rats is associated withhypophosphatemia, hypophosphaturia, hyperealcemia,and hypercalciuria (1-6). Because renal tubular trans-port of calcium may share a common reabsorptivemechanism or pathway with magnesium (7-9), onemight expect magnesuria with phosphate depletion.Data on renal handling of magnesium and its homeo-stasis during phosphate depletion are limited. Coburnand Massry (10) found that modest magnesuria withoutconsistent changes in plasma magnesium may occurduring phosphate depletion in adult dogs. Cuisinier-Gleizes et al. (6), reported a fall in the serumconcentration of magnesium and an increase in theurinary magnesium excretion in phosphate-depletedgrowing rats. The source of the urinary magnesiumlosses was not identified in their studies. Finally, dataon intestinal absorption of magnesium during phos-phate depletion are not available.

The present study was undertaken to evaluate theeffects of phosphate depletion on magnesium homeo-stasis with special emphasis on the renal handling,gastrointestinal absorption, and tissue content of mag-nesium.

METHODS

In preliminary studies in our laboratorv we found that rats fed alow phosphate diet do not ingest adequate amounts of food andtherefore do not gain weight as animals receiving a normal diet(Fig. 1). Because this phenomenon might affect variouisphysiological functions, the proper control group would be

The Journal of Clinical Investigation Volume 61 March 1978 573-581 573

350r

300F

2501BODYWEIGHT

(GRAMS)200

150-

100J

0 1 4 8TIME OF PHOSPHATEDEPLETION(WEEKS)

FIGURE 1 The effect of low dietary phosphate on bodyweight. Closed circles represent data from rats ingesting lowphosphate diet and open triangles denote data from ratsreceiving control diet. Brackets indicate 1 SE.

animiials receiving the control diet in an amount adjusted tomiiaintaini their weight e(qual to the phosphate-depleted group.

Male Sprague-Dawley rats 3-wk old were housed inindividual metabolic cages for balance studies. After a controlperiod of 5 days, the rats were allocated randomly and

sut)divided into two groups. The first received a lowphospholrus diet (0.03%) and will be referred to as phosphate-depleted (PD)' rats. The second group ingested a control dietconitainiing 0.44% phosphorus and were pair weighed to thePDgroup as described above; these animals will be referred toas pair-weighed (PW) rats. The dietary contenit of sodium(0.4%), calcium (0.41%), and magnesium (0.03%) were

identical in the control and low phosphate diets. All animalshad free access to deionized water.

Food intake, urinary and fecal excretions, and weight were

measured daily for the first 14 days and then at the last 2 days ofthe 3rd, 4th, 6th, and 8th wk of PD. Brilliant blue was used as a

stool marker. Blood was obtained in heparinized tubes durinigthe conitrol period and on days 1, 4, 7, 14, 21, 28, 42, and 56 ofPD. Someof the animals were sacrificed after 3 days and at theend of the 2nd, and 6th wk of PD for measurement of

magnesium content in the kidney, liver, skeletal muscle, andbone.

All fecal excretions of each balance period was ashed at650°C for 24 h in a muffle furnace and then extracted with0.75 N nitric acid for the measurements of magnesium.Samples of soft tissues were first weighed, then dried at 1050Cfor 48 h, defatted with ether, redried and weighed. The fat-freedry samples were ashed at 650°C for 24 h in a muffle furnace.The ash was extracted in 0.75 N nitric acid, agitated for 24 hand filtered for subsequent chemical analysis. Tibiae were

removed and completely cleaned of soft tissues; and shafts ofthe tibiae were split longitudinally and the bone marrov

' Abbreviations used in this paper: PD, phosphate deple-tioin(ed); PTH, parathyroid hormone; PW, pair-weighed rats.

removed. Samiiples of the bone were treated like soft tisstues asdescri)ed aI)ove.

To evaluate the possibility that chaniges in magnesiumhomiieostasis durinig PD mlay be related to the age of theanimals, studies were carried out in three additionial groups ofrats that were 6, 8, and 12-wk ol(l.

In aniother group, ninle rats fedc a low phosphorus diet hladmagniesiumi chloride supplemented for 6 wk in drinking waterthat conitainied 0.5% MgCI2. Balance studies were performed insix PD rats and six PWrats receivinig MgCl2 supplemenitationdurinig the 1st wk of the study.

To studv the effect of low dietary calciumii on the magnesuria,10 rats 8-wk old were fed a low phosphate diet in which thecalciuml- co'ntent was reduced to 0.01%.

To examiiine the effect of low dietary phosphate o01glomerular filtration rate, iniulin clearance was meeasured wZith[3H]inulin (New England Nuclear, Boston, Mass.) in 11i)ormiial rats, in 12 rats after 2 wk of PD, and in five animalsafter 8 wk of PD. The measuremiients were made in awake ratsrestratined in plastic cages. Radioactive inulin in 0.45% salinewas infuise(d in the jugular vein at a rate of 0.05-0.07 ml/minldeliverinig 0.25 ,uCi/min. After anl e(quilibration-i period of60-90nmill, turinie collections were obtained for three periodsof 15-30 min each; blood was obtained from tail vein at themi(lpoinit of eatch clearance period. The concentration of [3H]-intulin in blood anid urinie was cletermined with Beckmanili(px id scintillationi counlter, model LS-230 (Beckmana Inistrui-menits, Inic., Fullerton, Calif.).

Plasma inorganiic phosphorus was determinied by themicromethod of Chen et al. (11). Plasma, turine, fecal and tissuiemagnesitumii, and(I plaslla calciuimii were imieastured with thePerkini Elmer atomiiic absorption spectrophotometer. model503 (Perkin Elmer Corp., Norwalk, Conn.). Plasma and uirinesodiumiii were measuired by IL flame photometer (Instrumlielnta-tion Laboratory, Inic., Lexinigton, Mass.) whereas plasmia aiindurinarv creatininie were determined using a Techniconiautoanalyzer (Techniicoii Instruimenits Corp., Tarrytown,N. Y.). The statistical signiificance of the data was assessed bythe Studenit's t test.

RESULTS

After 2 wk of PD, the hair of maniy of the animals becamecoarse and patchy areas of hair loss appeared; theexposed skin was erythematouis. With prolonged PD,the rats appeared weak and sluggish.

The effects of PD on plasma concentrationis ofphosphortus anid mlagniesitum in rats 3-wk old are shownin Fig. 2. Plasma phosphorus fell significantly(P < 0.01) fromii 10.1+(0.27 (SE) to 5.0±0.54 mg/100within 1 day of feeding the rats a low phosphate diet.This was followed by a continiuous and gradual fall tothe lowest level of 1.2±0.21 mg/100 ml at the end of the8th wk of PD. The PWrats maintained their plasmaphosphorus within the normal range for the first 2 wk,but the level of phosphorus fell slightly but significantlyto 7.1+(0.24 mg/100 ml by the end of the 8th wk. Theconcentration of plasmla magnesitumii fell significantly(P < 0.01) from 1.2±.02 to 0.79±0.10 meq/liter duringthe 1st day of PD and reached its lowest level of0.48±0.01 meq/liter by the end of the 3rd wk. Theplasma magnesium levels in the PWrats did not change.The concentration of serum calcium was 10.2+0.11

574 W. J. Kreusser, K. Kurokawa, E. Azntar, E. Sachtjetn, and S. G. Massry

I- im - t

300[

URINARY MAGNESIUMEXCRETION 20C,ueq/24h

100[

0 4 7 2 3" 8

i - DAYS I EEKS- ITIME OF PHOSPHATEDEPLETION

FIGURE 2 Changes in the concentration of plasma mag-

nesium and phosphorus in PD rats (0) and pair-fed rats (0).Brackets denote 1 SE.

mg/100 ml before PD and 12.4±0.44, 12.4+0.20,12.0±0.36, 9.9+1.70, 10.4+0.21 mg/100 ml at 1, 2, 4, 6,and 8 wk of PD.

A significant (P < 0.01) increment in urinary mag-nesium excretion from 46±2.7 to 126+24 ,ueq/24 h was

noted during the 1st day of PD. Urinary magnesiumcontinued to increase and reached its peak of 300±24,ueq/24 h during the 3rd day anid remained highfluctuating between 150-300 ,eq/24 h thereafter (Fig.3). The magnitude of the magnesuria was related to thedegree of hypophosphatemia (Fig. 4); the correlationcoefficient for this relationship is 0.74 with P < 0.01.The PWrats did not display a substantial change inurinary magnesium. The PDrats also displayed markedhypercalciuria; urinary calcium increased from a

control value of 0.40±0.09 mg/24 h to 23.3±3.2 mg/24 h(P < 0.01) on the 3rd day of PD. This is in contrast to PW

URINARY

MAGNESIUM

EXCRETION /,Aeq/24h

100 _

O _ i, IX0 5 lo 14Y/3 #8

DAYS WEEKS---ITIME OF PHOSPHATEDEPLETION

FIGURE 3 The changes in urinary excretion of maginesium inrats with PD(0) and in pair-fed rats (0). Brackets denote 1 SE.

**

40.

.000**;0

0.

* 0

* .

* 0

o_n

OOD

Q0O

: e 'CBo° 8 ^¢

0040°O 009$X

0 2 4 6 8 10PLASMAPHOSPHORUS

mg/lOOmI

12

FIGURE 4 The relationship between the concentration ofphosphorus in plasma and urinary excretion of magnesium.Closed symbols denote data from rats ingesting a low dietaryphosphate and open svmbols represent data from pair-fedanimals.

rats that had no significant change in urinary calciumexcretion (from control value of 0.41+0.06 to 0.34+±0.06mg/24 h).

The concentration of serum creatinine did not changeduring PD; it was 0.42±0.02 mg/100 ml before PDand0.53±0.04, 0.50+0.05, 0.43+0.05, and 0.47±0.03mg/100 ml at 1, 2, 4, and 8 wk of PD respectively. The24 h excretion of creatinine was 2.1±0.16 mg/100 g

body wt per 24 h before PD and 2.8±0.38, 2.9±0.21,2.7±0.16, and 3.2±0.14 at 1, 2, 4, and 8 wk of PDindicating the adequacy of urinary collection . Meas-

urements of endogenouis creatinine clearance severaltimes throughout the study were not different from

values in PW*A rats. Furthermore, measuremiients ofglonmeruilar filtration rate with radioactive inulini re-

vealed no significant effect of PD. The glomerularfiltration rate was 1.27±0.01 ml/min per 100 g body wtin normal rats, 1.19±0.01 ml/min per 100 g body wtafter 2 wk of PD, and 1.00±0.15 ml/min per 100 g bodywt after 8 wk of PD. The blood levels of sodium did notchange and sodium excretion was not affected sig-nificantly by low dietary phosphorus. It was 1,388±163,ueq/24 h before PD and 1,330+218, 1,550+154, and1,598+274 ue(1/24 h at the end of the 1st, 3rd, anid 8thwk of PD, respectively.

The data on the balance of magnesiuim dturinig lowphosphorus intake are shown in Fig. 5. Despite thesignificant magnesiuria, fecal magniesium did notchange significantly dturing the first 2 vk of low dietaryphosphorus, and the fractioin of ingested magnesitumabsorbedl ranged between 0.48±0.05 and 0.61±0.04

Effect of Phosphate Depletioni on Magnesium Hoineostasis in Rats

1.0PLASMAMAGNESIUM

meq/liter 0.5

C10

PLASMAPHOSPHORUSmg/lOOmI 5

1.5r 400r

575

2001

_,) 100

LLJ

.4+mM 100n

ID 200

FIGzUIE(lit. ()ISolid1 bai

(luril) g(tiffe' i

ev'er', t

i nrera:

().69-3(]studv.duInril g

eI acli 11

Theand (ldawere', i

1)es inn11an(n PXbon)e ('(decrea:(01 tdl-)clplhosphconteill

(111 cl 4-'

cointent 0t bone in FLD rats was not (lltterei)t tronii thalt 0tPWVrats. The calciumnl to magniesilin) raitio in) I)one of PDrats was, therefore, signiificanitly higher than that in PW

fl ~~~ rats.

| .Kl1<>141m< XThe PD rats receiv iing MgC]2 in their drinkinig xvwatermiaintiainied 1)lood levels of migineiumis (1.25±t)0.04

_II II - me(-/liter and (lid not display hvponmagnesemia desliteldecremeints in the levels of serumn phosphor-us similar- tothose noted in PD rats without MgCl2 supplementation.

- ~ ~Despite a four- to fivefold increment in maKlgniesiumiiiinitatke by these r-ats, the co1ntent of magnesium in i)one

L ,, 1*emainiedi low andcl was 1)0t (lifferenit from tlaclt observed-2 0 4 8 14 3 4 6 8 in PD rats witlhouit MgCl suipplementtation (Table 1).

The balance (latta (duiring the lst w7k of the stu(lv in six5 \Magnesiimn l)balanc in raitts fed a low phosphlalte PD rats aniid six PW7rats receiving MgCLsupplemental-

I)e1 barsir il(Iic' urinailry (x'tiOlo tlalgnes%iumiit11(l lion are given in Table II. Althouigh the PD Iratts hiad ai.rs rel)res(nt feCcal magncsium. positivemgagniesini ml)alance, themagn.itude of the

positive balance wvas siglificantly (P < (.()1) lower thlanthis periocl. These valuecs are n)ot significantly in PWrats receiving the siame atmonint of mi.aginesiumiii

i)t froim0 the control values of ().55+0.03. How- suipplemiienitationi.the fl'actionl of inIgestecl imaitgnesiumiii absorbed The effects of low phosphorus intake on the levels ofsedl significantly (P < 0.01) to values of serum phosphorrus andl nmagnesitium and on urinary).06 aindi( 0.74±+(0.03 (luring the 3rd-8th wk of the exeretioni of milaginesiumil in okler ri-ats atre presented inThe balance of' miagnesiuim lbecame negative Table III. As in the younger rats, PD prodlucedl athe 1st day of low dietary plhosphorus acnd significant f'all in the coincenitrattionis of serumii phos-

e(l so throughout the enitire studl. phorus aind magniesiumiii (P < 0.01) aind ia significantresiults on miagnesitiuim contenit in variouis tissues incremilenit in urinary mi.agnesiuimil (P < 0.01) in raits 6, 8,,ta on bone calcitim arie given in) TIable 1. There and 12-wk ol0(. In these raits, the urinary excr-etioni ofi(o significainit (lifferelices letween the mtag- mlaginesitim before PDwas greater thanl in the youngeri conitenit of the mul1scle, kidney, ainld liver of' PD rats. This is at least partly duie to greater ainlouinits of f'oodV ratts. In conitriast, thew mialgnei4sitiliu content in ingeste( by the ol(ler rats. As in the case of the youngerlisplaved a modlest andcl significanit (P < 0.01) rats, the older animals hlad miarkedl hvpercalciu-ia.se withiin 3 days of foee dinlg wvith low dietary Urinary calciumiii inicreased 1b the 3rd day of P1) toiate. 'Ihe (liffereice l)etween the imiaginesiuimn 18.4±_2.2, 10.9± 2.2, anld 1 5.2 ± 3.5 in rats 6, 8, and 1 2-wkt of bone in) PD) and PXV rats becamse greater at 14 old respectively.

(dlas of the study (P < 0.01). The calciumiii The effect of low- dietary phosphate and(I calcium 0o

TABLE IEffj(4Nt *,f *, I)w Dic tfar Phosphorus mn Tissue Conitenit of Ala.giesiun7n anid on, Bone Magne.sinm tI nd Calcninm

Mlagnesillmll

GroIp I usele Liver Kidne\ B3on,e Bone calciuim ( alvi im'agnesi,

m(J11()0 g fat-free dril tet meqilOo( g 1iat-freef} (1r(/ ,tr

1 3 daysP1) 10.6±(0.52 7.9±'0.2:3 8.6±0.32 29.4±2.29*-- 1,334±25.2 49.1±+4.06)PW 10.9±.:36 7.1±(0.33 8.8±0.52 38.7± 1.45 1,385±_21.9 :35.9± 1.4-4

2 14 (lavsPI) 8.7(0.0:3 6.5±0.12 8.3±0.13 23.6 +(0.48* 1,138±35.2 48.1 ± 0.97P\V 9.8±(0.07 6.7±-0.11 8.7±-0.21 37.5±0.82 1,154+ 11.6 :31.()0±)0.45

3 42 da\ sP'1) 11.7±0:38 7.0±0.25 9.1±0.29 2)3.9± 0.9 1 1,183+:31.2 49.5± 1.19"PWV 11.7±+0.41 6.7±0.2:3 9.4±0.22 41.6±1.3(1 1,236±:30.5 29.8±X0.70P1) + \IgL 12.2±0.24 6.8±0.26 9.2±0.3:3 23.0 ±1.09*

I(lindatcs vlllles significatly (lifferent (P < 0.0(1) firo )PWratts.

576 XV. J. Kretisser, K. Kuriokawta. E. Az1nar, E. Sachtjen, (ai(I S. G. AlIa.ssrl

--. . 11 1- " 1--% I I'l, t, 1 V

I

I

3001

TABLE IIBalance Data in Six PD r(ats atnd Six PI' Rats Receiving

Alaginesinini Sutpplementationi duirinig theIst Xl7k of t/le Stuidil

Davs

2 aid 3 4 aniid 5 6 and(l 7

Mlagnesinm PD 1,164+57.(0 1,161±79.1 1,124-+-95.0initake, PW\ 1,175±81.2 1,180±69.8 1,164-58.0WX(1124 hi

Fecal inalg- I'D 411±47 .5 458±64.7 451±68.5n'esiuill), P\W7 426±51.() 463±27.() 449±67.8iLe(I 124 hi

Urinarv PD 493±36.1 * 471 +38.2* 485±46.1*nllaignie- P\W 350±20.9 3:30±2-7.0 340±32.1

W,r (/ I124 h1

MIagnesinm PD +260+36.5* +232±68.7* +188±88.1*b)alance, PW +399±35.3 +385±37.3 +375±41.0

cq 124 h

D)ata arie presented as \Ieai+±SEM\l.* Indicate value is signiificanitly- differenit (P < 0.01) fromii PW7rats.

urinary mlagnesiumiii anid calciumiii are givenl in Table IV.In these animals, urinary miiagiuesitun)i also iniereased tomlore than 300 me(I/24 h but the hypercalciuria wasmaslrkedly! ani(d sign ificaintlv (<0.01) less than in PD ratsreceiving higlher calcitumii initake. Calciumil excretionimicr-easeol froml 0.06±+(0.03 to 1.17±+(0.18 mg/24 h.

1DISCUSSI ON

The restults of'the present stuldy dlemiionistrate that PDini the r-at is associated with significanit inieremiienits inthe urinary excretory rates of magnesium, signiificanit(lecremiienits in the conicenitrationi of miiagniesitumii in theblood, and(1 negative imiagnesitm l)balance. Other inivesti-gators halve reported variable degrees of mnagnesuiriadturinig PD ini rats (6) (logs (10), anid main (12).

Ani increase in urinary miagnDesiunm may followel)hanceed initestinial absorption of this ionl, ani inicreasein its filteredl load, autgmeneted release of magnesiumfromii bone anid (or) soft tissues, a olecrease in its renialtubular reabsorption, or secretioni of magniesium by thenephron. Ou-r dlatta does not support the first twop0SSi)bilities. The inlcrease in urinary milagimesitlilloccurre(l before any evidlent change in its initestinalalsorption andc at a timiie wheni the 10lood levels ofmiagnesiumiii fell, aniid consequently filtered loads ofmiaulgniesiuim Xwere lower thain control levels.

Magnesitum conitenit of bone xv%,as dlecrieasedI by\ theend(I of 3 dlays of' feedinig the ralts a low phosplhate (lietandici remaiiine(d lo' tlhrouiglhouit the stdly. This wx'as notdute to an overall reduction in minieral-l content of' bone,

because changes in calcium contenit of bone were notdetected durinig PD. At least two possibilities couldaccount for the fall in bone magnesium. First, PD isassociated with the formatioin of magnesium-poor bonein these growiing rats; and second, PD may enhancemagniesiumii release from bone. Our data suggest that1)oth of these melclhanisms may be operative. Thefindings that rats with PD developed negative mag-nesiumii balance despite similar dietary intake ofmagniesiumiii and in the f:ace of no decrease or even aninicrease in the fractioni of ingested magnesium ab-sorb)ed favors loss of magnesium from body stores.Because only bone displayed a decrease in its mag-nesiumil conitenit, it is reasonable to assumiie that PD isassociatedl with an augmiiented release of magnesiumfromii the skeletoni. However, the observation that bonecontent of mnagnesiumll remainied low despite positivemagnesium 1balance in PD rats receiving magnesiumsupplenmentation is consistent with the formation ofmlagniesiumil-poor bone. The inability to increase bonemagnesiunm to normal despite positive magnesiumbalance is not surprising becaause the net balance is stillnegative when compared to PWV rats receiving similaramounits of' nagnesium supplemiienitation.

The magnesium loss from bone may be the primaryevenit in the genesis of the magnesuria but it may also besecondary to renal maginesium leak. It is welldocumiiienited that a normally funcetioninlg kidney has atremiieindous magnesium conserving ability (13-15),anid in the presenice of hypomagnesemia not due toprimary renal loss, magnesiumii virtually disappearsfrom the urine (13-15). The demonistration in thepresenit study that marked magnesuria persisted de-spite the hvpomagnesemia favors the postulate that PDinidtices renial mlaginesiumii wasting with the bone beingthe souirce of the urinary magnesiumil losses. It is ofinterest that the anmounit of magniesium that is requiredto prevent the ftall in its blood levels is very smiiall(15-20 /xeq) comlpared to the increimients in urinarymlagniesiumii. This quantitative disparity and the obser-vationi that levels of blood magnesiunm were maintainedat normal values only after the supplementation of verylarge quantities of' magniesiumii (800-1,000 ,ueq/day)provides further support for a primary renal leak ofmaginesiummi.

Other factors beside urinaryi magnesium losses maycontribute to the hypomagnesemia in our PD growingrats. It is interestiing that PD in humanis is associatedwith mlagnesiuria but wvith only' minor chainges in thelevels of' serumii mlEagInesiuim (12), anid in ouir PD adultrats, the fall in the coneenitrationi of serum magnesiumiiwas less thalnl in the xyounger rats despite similar or evengreater mlagniesuria. It is possible that the need forImlagiuemii by newvlv forimied bonle in the PD growingralts, wvhich are in negative mlaginesiuinm balance, coIn-tributes to the magnitude of the hypomagnesemia.

Effect of Phos'pha te Depletion oni Alagnesium Homieostasis in Rats 577

C5o0-+1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1~ -00 000-1000 -0In0 0

Cl 0" 0 - in ino 6 Cl 01 - Cl CO o

_ +l +l +1 +1 +l

<Oar +lCO+l -lX~~~~~~~~~~~~~~~~~~~~~~~~~t> t-

- ~0 +1 +1 +1

*> ~ ~~~oo 0zlF.s o 0 0 - 01C t

~̂Xt- +l +l + +1 +1

S~~~~~~ 0o00.bNN t-S

~ 0 00 CO

~~~0 0 CO~~~~~~~~~0O Cl C00 C

-+1 +1 +1 +1 +1 +1 +1 +1 +1_- zC0 ) 0 0

X°SYN CCO0_EODK > ° >+°+1 +1 +1

ot K O t -Oc00 00

<O t UrCO0000 C

001 CC000 +1 +1 +1

00 Cl~~~~0 00 CO_-0n 0-_c$ 0- +1 +1 +1 +1 +1 +1 +1 +1 +1 - +1 +1 +1

X~ ~ ~ 0o0 o 0- 00ci _ CO_ 0 00_0-_

0t 0 CO 00 - 10b V.N000~~~co 000~~~~~~~~~ 600qC 00 C

00 COCO

LO CIO00 ~~~~~ ~~~~~~~~~~+1+1 +1

- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~CO

co) ~ cor co,X U)*C C 000°; Da

C.IqQ CO

+1 +1 +1 +1 +1 +1 +1 +i +1

10 CO~~~~~~~~~~~~ q '

"~~~~~~~0 +1~~~~~~~~~~~~~~~~~+1 +16.a ~~~~~00-

CO C 00C CO

Z~~ COeO 0-0 0 10

+1 +1 +1 +1 +1 +1 +1 +1 +1C) 00 CO m 00 C,000 00 ~ CO - CO

OD~ ~ ~ ~ 0COcoCO

00 ~~~~~~000 -cq

C,)00co 000~~~~~~~~~~~~~~~1 +1 +1 65

00 00 vv~~~~~~~~~0

o - ~~~~~~~~~~~~~~~~~~~+ + 1+1+1 +1 +1 +1- 0- ~~~~~~~~~COCO! v CO C - 0

-00 0- 001 co 'o00 00 00 0 10 0 00

0 ++1 +1 +1 +1 +1 +1 -1 +1 +1 +1 o

00 lf~ 4 1 0 10 00 - 00- 0 10C')0

C Cli C-4~~~0zr.- ~ -01 ~ Cl

t6 t6k

00 CO~~~~~~~bo <C~~~~~~~~~~~~

578 W.J. Kreusser, K. Kurokawa, E. Aznar, E. Sachtjen, and S. G. Massry

TABLE IVEffect of Low Dietary Phosphorus acnd Calcium otn the Conicenttrationl of Serum Inorgacnic Phosphorus,

Magniesium and Calcium and Urinary Excretion of Magniesium and Calcium

Dass after low phosphate and calcitum diet

Control 1 2 3 4 5 6 7

Weight, g 266+5 275+±5Age, wk 8 9Serum P,

mg/i 00 ml 9.2+0.24 6.4+0.27 6.4±0.29 5.1+0.23Serum Mg,

meqlliter 1.70+(0.05 1.44±0.05 1.2+±0.03 1.2±0.04Urinary Mg,

peq124 h 200±10 210+21 275±22 300±26 320±19 320±+18 317±19 333+18Serum Ca,

mg/lO ml 10.3+(0.08 10.0±0.12 10.1+±0.09 10.2+0.14Urinary Ca,

mg/24 h 0.60+0.03 0.65+0.10 0.93±0.14 1.17±0.18 1.36±0.25 1.38+(0.26 1.3±0.20 2.5+0.62

Data are presented as mean±+SE.Abbreviations: P, phosphorus; Mg, magn?sium; Ca, calcium.

Our data indicate that a decrease in the net tubularreabsorption of magnesium occurs during PD becausethe magnesiuria persisted despite a decrease in filteredloads of magnesium. The mechanisms responsible forthis phenomena are not clear. Several possibilitiesshould be considered. First, this defect is obviously notrelated to the age or the weight of the rats because itoccurred in animals of different ages and weights.Secondl, a reduction in the tubular reabsorption ofmagnesium usually occurs with conditions causingnatriuresis (15-18). In our animals, magnesiuria de-veloped in the absence of changes in urinary sodiumexcretion.

Third, magnesiuria occurs in animals (19, 20), orhumans (21) under the influence of chronic min-eralocorticoid excess. It is possible that PD inducesthe release of steroids with mineralocorticoid proper-ties, although this possibility seems remote. Animalstudies have shown that with initiation of mineralcor-ticoid administration, there is a period of sodiumretention without magnesiuria, but urinary magnesiumexcretion increases as the animals escape the sodiumretaining effect of the hormone (20). In our study,magnesuria occurred during the 1st day of PDand therewas no evidence of sodium retention.

Fourth, PDhas been reported to cause hypofunctionof the parathyoid glands (12, 22) and becauseparathyroid hormone (PTH) enhances magnesiumreabsorption (23, 24), partial or complete lack of PTHcould cause modest magnesuria. However, the data ofCuisinier-Gleizes et al. (7) argues against this possibil-ity. They found that magnesuria occurred inthyroparathyroidectomized PD rats. Also, availabledata indicate a resistance to the phosphaturic action

of PTH in PD rats (25). Theoretically, a resistanceto the action of PTH on renal handling of magnesiummay develop during PD and as such contribute to themagnesuria.

Fifth, mild elevation in the blood levels of calciumiidevelop during PD (1, 6, 10, 26), and this may be, atleast, partly responsible for the magnesuria becausehypercalcemia decreases the renal tubular reabsorptionof magnesium (9). Indeed, the concentration of bloodcalcium in our rats increased by 1.0-2.0 mg/100 mlduring the first 4 wks of PD. However, the blood levelsof calcium returned to normal between the 5th and 8thwk of PD but the magnesuria persisted. Theseobservations clearly indicate the renal magnesiumwastiing during PD could not be entirely explained bythe changes in the concentration of calcium in theblood. Also, hypercalciuria of PDmay play a role in thegenesis of the magnesiuria. However, this seemsunlikely because marked magnesiuria persisted whenthe hypercalciuria was markedly blunted in rats fed alow phosphate and calcium diet.

Finally, preliminary data by Ben-Isaac et al. (27)showed that the blood of PD rabbits contains a humoralfactor which suppresses the tubular reabsorption ofcalcium. It is well documented that the renal handlingof calcium and magnesium is closely related (7-9), andit is reasonable to assume that such a humoral factor mayalso reduce the tubular transport of magnesium.

A decrease in the net tubular reabsorption ofmagnesium during PD is not surprising because othershave reported tubular reabsorptive defects of othersubstances such as calcium (10), bicarbonate (28),glucose (29), and sodium (30).

One should also consider the possibility that PDmay

Effect of Phosphate Depletion on Magnesium Homeostasis in Rats 579

be associated with renal tubular secretion of mag-nesium. The evidence for the existence of magnesiumsecretion by the renal tubule is conflicting. AlthoughAverill and Heaton (31) demonstrated magnesiumsecretion in the rat during magnesium loading, Alfred-son and Walser (32) failed to confirm this observation.Recent micropuncture data obtained in magnesium-loaded rats suggest that magnesium may be secreted bythe descending limb of the loop of Henle (33). In thedog, Massry et al. (24) were unable to show tubularsecretion of magnesium although others found thaturinary magnesium may reach values that are 10-20%greater than filtered magnesium durinig the concomi-tant admipistration of magnesium salts, saline, andfurosemide (34). It appears that if magnesium secretionby the nephron exists, it plays a minor role in the renalhandling of magnesium.

It is evident that PDcan induce marked alteration inmagnesium homeostasis. Our data permit the formula-tion of a certain sequence of events in the relationshipbetween body stores of phosphate and magnesiummetabolism. It appears that PD induces, by as yet,undertermined mechanism(s) a decrease in the tubularreabsorption of magnesium and results in magnesuria.The enhanced urinary losses of this ion are followed byhypomagnesemia and negative magnesium balancewith bone being the main organ affected. An adaptiveaugmentation in the fractional intestinal absorption ofmagnesium occurred only after a prolonged period ofPD and when the cumulative negative balance ofmagnesium was marked.

ACKNOWLEDGMENTS

The authors wish to thank Mrs. Alice Peterson, Gracy Fick,and Thelma Alvarado for their secretarial assistance, and Ms.Virginia Berberian and Mr. Barry Gammel for their technicalhelp.

This work was supported in part by a grant from the KidneyFoundation of Southern California.

REFERENCES

1. Shikita, M. S., S. Tsurufuji, and Y. Ito. 1962. Adaptation inrenal phosphorus excretion under the influence ofparathyroids; a study of unilaterally catheterized rats.Endocrinol. Jpn. 9: 171-180.

2. Steele, T. H., J. E. Engle, Y. Tanaka, R. S. Lorene, K. L.Dudgeson, and H. F. DeLuca. 1975. On the phosphaturicaction of 1,25 dihydroxvvitamin D3. Am. J. Physiol. 229:489-495.

3. Steele, T. H., and H. F. DeLuca. 1976. Influence of dietarvphosphate on renal phosphate reabsorption in theparathyroidectomized rat. J. Clin. Invest. 57: 867-874.

4. Tr6ehler, U., J-P. Bonjour, and H. Fleisch. 1976. Inorganicphosphate homeostasis. Renal adaptation to the dietaryintake in intact and thyroparathyroidectomized rats. J.Clin. Invest. 57: 264-273.

5. Day, H. G., and E. V. McCallum. 1939. Mineralmetabolism, growth, and symptomnatology of rats on a diet

extremely deficient in phosphorus. J. Biol. Chemii. 130:269-283.

6. Cuisinier-Gleizes, P., M. Thomasset, F. Saintenv-Debove, and H. Niathieu. 1976. Phosphorus deficiency,parathyroid hormone, and bone resorption in the growingrat. Calcif. Tissue Res. 20: 235-249.

7. Samiy, A. H. E., J. L. Brown, D. L. Globus, R. H. Kessler,and D. D. Thompson. 1960. Interrelation between therenal transport systems of magniesiumll and calciumiii. Ami. J.Physiol. 198: 599-602.

8. Walser, M. 1967. NMagnesium metabolism. Ergeb. Philsiol.Biol. Chem. Ex?). Pharmakol. 59: 185-296.

9. Coburn, J. W., S. G. Massry, and C. R. Kleeman. 1970. Theeffect of calcium infusioni on renal handlinig of magnesiuimwith normal and reduced glomerular filtration rate.Nephron. 7: 131-143.

10. Coburn, J. XV., and S. G. Massrv. 1970. Chaniges in serumand urinary calcium during phosphate depletion: Studieson NMechanisms.J. Clitn. Itnvest. 49: 1073-1087.

11. Chen, P. S., Jr., T. Y. Toribara, and H. Warner. 1956.Microdeterminationi of phosphorus. Atnal. Chem1. 28:1756-1758.

12. Dominguez, J. H., R. W. Gray, and J. Lemanni, Jr. 1976.Dietary phosphate deprivation in womenand men: effectson mineral and acid balances, parathyroid hormone andthe metabolism of 25-OH-Vitamin D.J. Clin. Enidocrinol.Metab. 43: 1056-1068.

13. Fitzgerald, M. G., and P. Fourman. 1956. An experimentalstudy of magnesiuim deficiency in man. Clini. Sci. 15:635-647.

14. Barnes, B. A., 0. Cope, and T. Harrison. 1958. Magnesiumconservationi in the human being on a low magniesiumdiet.J. Clin. Inivest. 37: 430-440.

15. Massry, S. G., and J. XV. Coburn. 1973. The hormonial anidnon-hormonal control of renal exeretion of calciumiii andmagnesium. Nephron. 10: 66-112.

16. Massry, S. G., J. W. Coburn, L. XV. Chapman, and C. R.Kleeman. 1967. Effect of NaCl infusioni on urinary Ca"tand MgX++ during reduction in their filtered loads. Amii. J.Phijsiol. 213: 1218-1224.

17. Eknoyan, G., V. N. Suki, andl M. Mlartinez-Maldonado.1970. Effect of diuretics on urinary exeretioni of phosplhate,calcium and magnesium in thyroparathyroidectoimnizeddogs. J. Lab. Clitn. Med. 76: 257-266.

18. Ahumada, J. J. anid S. G. Massry. 1971. Renial vasodilata-tion: Effect on renal handling of phosphate. Clin. Sci.(Oxf). 41: 109-121.

19. Scott, D., and A. Dobson. 1965. Aldosterone and themetabolism of magnesium and other minerals in sheep. Q.

J. Exp. Phiysiol. Cogni. Med. Sci. 50: 42-56.20. Massry, S. G., J. W. Coburn, L. W. Chapman, and C. R.

Kleeman. 1968. The effect of long term desoxycortico-sterone acetate administration on the renal excretion ofcalcium and magnesium.J. Lab. Clin. Med. 71: 212-219.

21. Hortoni, R., and E. G. Biglieri. 1962. Effect of aldosteroneon the metabolism of magnesium. J. Clitn. Enidocrinol.Metab. 22: 1187-1192.

22. Stoerk, H. C., and W. H. Carnes. 1945. The relatioin of thedietary Ca:P ratio to seruim Ca and to parathyroid volume.

J. Nutr. 29: 43-50.23. Bethune, J. E., R. A. Turpin, and H. Inoutie. 1968. Effect of

parathvroid extract on divalent ioni exeretioin in man. J.Clit. Endocrinol. Metab. 28: 673-678.

24. Massrv. S. G., J. W. Coburn, and C. R. Kleemiiai. 1969.Renal hancdling of magnesitumii in the dog. Amii. J. Phi.siol.216: 1460-1467.

25. Steele, T. fI. 1976. Renal resistanice to parathyroid

580 W. J. Kreusser, K. Kurokawa, E. Aznar, E. Sachtjen, and S. G. Massry

hormone during phosphorus deprivation. J. Clin. Invest.58: 1461-1464.

26. Freeman, S., and F. C. MIcLean. 1941. Experimentalrickets. Blood and tissue changes in puppies receivinig adiet very low in phosphorus with and without vitamin D.Arc/i. Pathol. 32: 387-408.

27. Ben-Isaac, C., S. G. Massry, S. Rosenfeld. C. R. Kleeman,and NM. Bick. 1974. Evidence for humoral factor responsi-ble for the hypercalciuria of phosphate depletion.J. Clini.Inlvest. 53: 5a. (Abstr.)

28. Gold, L. NI., S. G. Massry, A. I. Arieff, and J. W'. Coburn.1973. Renal bicarbonate wasting during phosphate deple-tion. A possible cause of altered acid-base homeostasis inhyperparathyroidism. J. Clin. Inivest. 52: 2556-2562.

29. Gold, L. MI., S. G. Nlassrv, and R. NI. Friedler. 1977. Effect

of phosphate depletion on renal tul)ular reabsorption ofglucose.J. Lab. Clini. Med. 89: 554-559.

30. Goldfarb, S., G. R. Westby, MI. Goldberg, and Z. S. Agtus.1977. Renal tubular effects of chronie phosphate deple-tion.J. Clini. Invest. 59: 770-779.

31. Averill, C. NI., and F. W. Heaton. 1960. The renal handlingof magniesium. Clini. Sci. (Oxf). 31: 353-360.

32. Alfredson, K. S., and NI. Walser. 1970. Is magnesiumsecreted by the rat renal tubule? Nep/hron. 7: 241-247.

33. Brunette, NI. G., N. V7igneault, and S. Cariere. 197'5.Mlicropuncture study of renal magnesium transport inmagnesium-loaded rats. Ami.J. Pliysiol. 229: 1695-1701.

34. Wen, S. F., N. L. Wong, and J. H. Dirks. 1971. Evideniee forrencal magnesium secretion during magnesium infuision inthe dog. Amii J. PhyJsiol. 220: 33-37.

Effect of Phosphate Depletion oWAloMagnesium Homeostasis in Rats 581